Monday, July 21, 2008

Kai-Fu Lee keynote at SIGIR

Googler Kai-Fu Lee gave a keynote talk yesterday at SIGIR 2008 on "The Google China Experience".

The Google China experience has been fraught with difficulties. As Kai-Fu said, Google was expecting that their search engine could succeed in China with the same interface and approach that had worked so well elsewhere, but was "humbled" by the lack of uptake.

Why did a Google that had worked well elsewhere not succeed in China?

Kai-Fu argued that two unusual features of Chinese internet users, their youth and their language, appears to explain the difference.

While showing examples of sites that did succeed in China, Kai-Fu argued that the young, novice, hurried average Chinese user is looking less for navigation or one authoritative sources of information and more for entertainment and broad surveys of information.

Google China was optimized for finding the one site you need to go to, as it is elsewhere, but, Kai-Fu said, according to eyetracking studies and log data, Chinese users tend to be much less task-oriented, read much more of the page, and click many more links than US users.

So, Google started to offer up more opportunities for exploration and discovery -- by being much more aggressive with query suggestions and universal search, for example, and by adding browse links to the Google front page rather than just having a stark page with just a search box -- rather than just trying to give a quick answer and let people move on.

Kai-Fu suggested that the Chinese interest in browsing and exploration also may have roots in the Chinese language itself. Chinese is slower and more work to type than other languages, but more compact to read. This seems to cause a preference for clicking over typing and a preference for pages dense with information over the sparsity Google tended to favor in other countries.

One curious question that Kai-Fu raised was whether these preferences will remain true over time. Expert internet users tend to be more task-oriented than novice users. Google China has had much more success in gaining market share in China among expert users.

It may be the case that, as people gain more experience with the Web and what is available on the Web, their behavior shifts from browse to search, from exploration of what is out there to finding the information they know must exist.

Update: Paul Heymann at Stanford Infolab posts an excellent summary of Kai-Fu Lee's talk, as well as notes on other talks at SIGIR.


Anonymous said...

While showing examples of sites that did succeed in China, Kai-Fu argued that the young, novice, hurried average Chinese user is looking less for navigation or one authoritative sources of information and more for entertainment and broad surveys of information.

hmm.. I don't remember Kai-fu saying that the average young novice user was "hurried". But yes, I found it fascinating that chinese users want more information, not less. They want exploratory interfaces, wherein they are made aware of what other options are available, rather than barren interfaces, with barely enough information to know how to take the next step in one's search task.


It may be the case that, as people gain more experience with the Web and what is available on the Web, their behavior shifts from browse to search, from exploration of what is out there to finding the information they know must exist.

I don't quite agree with this dichotomy. Switching from "browse" to "search" isn't what is happening here. Kai-fu did talk about chinese users in search mode, not in browse mode. And in search mode, they still do want to see more query suggestions, related items, etc.

I think there is a certain value to seeing what the range of available information is, for one of your searches, even if you never end up clicking on most of the suggested refinements. Maybe chinese users feel the same way? Maybe there is a hunger or thirst.. not for entertainment, but for new knowledge and understanding, with a wide-open curiosity for the world around them.

Whatever the reasons, I also found it fascinating to see the differences between wide user bases.

Greg Linden said...

Hi, Jeremy. Just briefly, on the "hurried", I was trying to summarize what Kai-Fu said about how Chinese users are often in internet cafes and being charged by the minute, so they are looking to do things now and fast.

Anonymous said...

Oh, yes. My bad -- apologies. He did indeed say that. I get what you mean now, and agree.

But, if that's the case.. if they are so hurried.. wouldn't they be more focused on the quick, no-fuss, no-muss, get in and get out U.S. style of Googling? It actually costs them money to be more "exploratory" in their searching habits. And yet it is valuable enough to them, to have all this additional information, that they still want it anyway.

It's really fascinating to me.

Anonymous said...

Nice speaking with you at SIGIR, though briefly. My notes on the keynote is now up -- Kai-Fu @ SIGIR.

Anonymous said...

pranam: Nice summary.. but.. what are your thoughts? What are you opinions on some of his conclusions? Do you agree with what Kai-Fu says? Disagree?

Take a statement like this: "Since users pay by the hour in Internet Cafes, they love the concept of directories."

Do you believe it? Do you not believe it? Why? To me, it sounds very odd to say that pay-by-the-hour implies love-of-directories. Isn't the whole point behind Google search the idea that you don't want to spend a lot of time finding the information that you need? Get on the search engine, type 1.6 words, look at the top 3 links, and then get off the search engine just as quickly? It seems that if you are pressed for time, and every second cost you more money, you would follow that model. But instead, Chinese users spend 30-60 seconds looking at search results, rather than the 10 seconds spent by U.S. users. Is that really as counterintuitive as I believe it is? Is there some hidden factor at work here? Or is it completely consistent?

I did a quick poll around my research lab back here in the U.S., and found that my Chinese-native co-workers also approach search in a different manner, even when using the Internet here in the U.S. My survey is only anecdotal at this point, but it suggests that the "pay-by-the-hour" Internet cafe explanation has nothing to do with it. Nor does Kai-Fu's explanation that the youth, inexperience, and naivete of the Chinese users cause this difference, because my coworkers are Chinese PhD researchers -- folks that are not young or naive or inexperienced. And yet they still "browse" the search results in a different manner than their U.S. counterparts.

So what are your own thoughts/feelings/ideas/opinions about this information? Are Chinese users really different? Do they really prefer more information than U.S. users? Is there something cultural happening here? Or is there something conditional happening.. maybe we've trained ourselves to think only in certain ways?

I'd be interested in getting a discussion going.

Anonymous said...


So much for Web 2.0 as a conversation medium :-|

Greg Linden said...

Hi, Jeremy. Yeah, sorry, I don't know enough about China to be able to comment intelligently. Perhaps someone with area expertise could chime in.

Anonymous said...

Well, I don't think either of us have to know much about China to have the meta-discussion that arises from this Kai-fu information.

What I am trying to say is that Kai-fu is telling us that various users differ not only in their estimates of relevance, but also in the strategies that each user employs in his or her information seeking behavior.

So, for example, think about two U.S. users, one that likes Jaguar the Mac OS X, and one that likes Jaguar the car. You would personalize each person's results, so that the ranked list would be biased toward one or the other, depending on the user model, correct?

Nevertheless, in both cases, if you were using Google, the user would see the same exact interface: 2-3 ads at the top, followed by 3-4 results above the fold, with 4-5 more ads on the right.

Even though the results themselves are different, there would be no tailoring of the interface to each person's varied information seeking behavior. There would be personalization of content, but not of style or method or interaction.

What Kai-fu found was that Chinese users differed significantly from their U.S. counterparts in term of information seeking behavior. The Chinese users liked to look at a lot more information in conjunction with their queries, and relied a lot more on link-based navigation, rather than multiple-query retyping, as a way of interacting with the search engine (as a way of refining their queries to meet their ongoing information needs).

What that says to me is that it is not just the ordering of the results that need to be personalized. It says to me that the interaction paradigm itself, in terms of how one carries on a conversation with a search engine, needs to be personalized. Different people have different modes, different styles, for communicating their information needs to a search engine.

To me, that has broader implications than just U.S. vs. China. To me, that says that different users have different methods and modes of search engine interaction. And that the search engines should not force a user into the search engine's only interaction mode. The mode itself should be as personalized or personalizable as the content of the results.

And yet, no search engine does this right now. Every major web search engine has a single interaction mode, and forces every user to adopt that mode and only that mode.

It's high time we expand the scope of what personalization means, to include one's information seeking strategy, in addition to disambiguatory content like jaguar/Jaguar! :-)

Anonymous said...

BTW, Greg, I didn't mean to imply that you had to respond. I was hoping that Pranam would respond. Get the web 2.0 all-around conversation going, instead of just me and you :-)

ETA said...

I'm a native Chinese user, so I agree most of what Kai-Fu said based on my own experience and observation.
First of all, English term is unique and widely accepted. Though chinese word is same accurate, but chinese words can be expressed in more way than English. So you can't rely on one search to get to what you want.
And on the other side, because chinese words are more concise, many user are able to filter out many unwanted pages simply by check the preview paragraph in the search page. And that also explained the 'hurried user'. It's actually 'quicker' user.
As for myself, I set my preference for search result per page to 50~100, rather than the default 10 result.
Secondly, many young Chinese user doesn't know how to optimize their search by using advanced search syntax like 'NOT','OR','AND', this lead to more search time.
And I also use English google search at work, and I found I have to dig into the page to see if the content match my expectation, which also lead to less search times. But on the contrary, the quality of the content is higher. I guess it's due to Google give higher priority to some famous sites, whereas China doesn't have many world-renown sites.
What's more, it is not true that Internet Cafe user want to stay long at search page....I have to say that's because many Chinese articles are copied here and there, so you have to roam around the search page many times to find some original articles.